Cognitive-behavioral therapy (CBT) constitutes an empirically based treatment for tic disorders (TD), but much remains to be learned about its impact at the neural level. Therefore, we examined the electrophysiological correlates of CBT in TD patients, and we evaluated the utility of event-related potentials (ERP) as predictors of CBT outcome. ERPs were recorded during a stimulus-response compatibility (SRC) task in 26 TD patients and 26 healthy controls. Recordings were performed twice, before and after CBT in TD patients, and with a similar time interval in healthy controls. The stimulus- and response-locked lateralized readiness potentials (sLRP & rLRP) were assessed, as well as the N200 and the P300. The results revealed that before CBT, TD patients showed a delayed sLRP onset and larger amplitude of both the sLRP and rLRP peaks, in comparison with healthy controls. The CBT induced an acceleration of the sLRP onset and a reduction of the rLRP peak amplitude. Compared to healthy controls, TD patients showed a more frontal distribution of the No-Go P300, which was however not affected by CBT. Finally, a multiple linear regression analysis including the N200 and the incompatible sLRP onset corroborated a predictive model of therapeutic outcome, which explained 43% of the variance in tic reduction following CBT. The current study provided evidence that CBT can selectively normalize motor processes relative to stimulus-response compatibility in TD patients. Also, ERPs can predict the amount of tic symptoms improvement induced by the CBT and might therefore improve treatment modality allocation among TD patients.